Case Analysis Mining Based On Cluster Analysis And Decision Tree Algorithm

In recent years,with the constant deepening of modern policing work,the construction of police policing information has also rapidly advanced.More and more public security police information is collected and collected by public security police of all police types,and the use of public security police information system is more and more frequent,so that the public security work is gradually transformed from the traditional static management into dynamic management and pay more attention to " Anti-based,fight anti-combination " approach.In the current information society,who have more information,who have the initiative to work.Therefore,establishing a scientific and efficient information working mechanism,timely and comprehensively grasping all kinds of public order information,especially deep-seated and early-warning information,and eliminating all kinds of hidden dangers of public order so as to truly guard against any possible precautionary measures and become the current public security The inevitable requirement of work.The public security department has accumulated a large number of cases of physical evidence,and the relevant departments can only make basic enquiries,the use of case material information has only one function..This paper first analyzes the characteristics of the case evidence information,use cleansing,integration,transformation,and other methods to process data and normalize the data.Then this paper analyzes the methods that can be used for case information mining to illustrate the possibility of data mining in case analysis applications.Using the characteristics like the case types in the case information,the geographical location of the case and so on,establish a corresponding model.Using k-means algorithm’s simplicity,high efficiency,and easy implementation to predict the probability of different cases in different regions,and use the hierarchical clustering algorithm to predict the shortcomings of the k-means algorithm.Then,in the defects and arguments of the traditional k-means algorithm and hierarchical clustering algorithm,improve algorithm.Through experiments and comparison with each other,hierarchical clustering improved algorithms can more accurately classify cases.In order to determine the importance of various attributes to the classification of the case.According to the characteristics of the case information,the ID3 algorithm is used for case prediction,provide additional information to related departments,facilitate the relevant authorities to effectively fight crimes.Finally,experimental data show that,the improved algorithm for hierarchical clustering algorithm and ID3 algorithm in this paper has achieved more accurate results and has certain practical significance.